MITRE ATT&CK technique · T1611
Escape to Host
How does container escape to host work, and why does it matter in AWS? Adversaries break out of a container to the underlying host, gaining access to other containers and the host’s credentials — in AWS, typically the node’s IAM role.
For leadership
Executive takeaway
How it works in AWS
- Abuse privileged mode, host namespaces or hostPath mounts.
- Exploit kernel or runtime vulnerabilities.
- From the host, request node credentials from the metadata service.
What the attacker needs
- Code execution in a container
- Privileged settings or exploitable vulnerability
Relevant AWS API calls
| API / event | Why it matters |
|---|---|
| IMDS /latest/meta-data/iam/security-credentials/ | Node role credentials after escape (not a logged API call). |
A risky permission is not an observed attack
Privileged pods, hostPath mounts, IMDS hop limit > 1, broad node roles.
Runtime findings for escape behaviour; node role used for unexpected API calls.
Preventive and limiting controls
- Restricted Pod Security Standard
- IMDSv2 hop limit 1
- Minimal node roles and per-pod identities
Mapped cloud controls
- preventiveRestricted pod security settings
Blocking privileged containers, host namespaces and hostPath mounts removes the most common escape routes.
- restrictiveEnforce IMDSv2 and limit metadata reachability
With IMDSv2 required, a hop limit of 1 stops the session-token response from reaching containers, so they cannot use the node metadata service and its credentials.
- detectiveManaged threat detection across accounts and regions
GuardDuty Runtime Monitoring can surface container escape and privilege escalation behaviour on nodes.
See assumptions and validation evidence in the Cloud Control Explorer →
Detection signals and telemetry
- Escape and privilege-escalation behaviour — GuardDuty Runtime Monitoring
- Node role performing write actions — CloudTrail
MITRE detection strategies (official, cloud-relevant analytics)
- DET0219 Detection Strategy for Escape to Host
- AN0612 (Containers): Detection of container escape attempts via bind mounts, privileged containers, or abuse of docker.sock. Defenders may observe anomalous volume mount configurations (e.g., hostPath to / or /proc), unexpected privileged container launches, or use of container administration commands to access host resources.Log sources: docker:daemon — container create/start with privileged flag or host volume mount; kubernetes:apiserver — Pod spec with hostPath or privileged securityContext
MITRE mitigations
Official MITRE description (excerpt)
Adversaries may break out of a container or virtualized environment to gain access to the underlying host. This can allow an adversary access to other containerized or virtualized resources from the host level or to the host itself. In principle, containerized / virtualized resources should provide a clear separation of application functionality and be isolated from the host environment.
Common misconceptions
- “Containers are a security boundary like VMs.”
- Containers share the kernel; misconfiguration makes escape straightforward.
AI security
Related MITRE ATLAS techniques
ATLAS and ATT&CK are separate frameworks. These relationships are listed explicitly, with their basis.
- AML.T0105 Escape to HostPublished by ATLAS as its ATT&CK reference — ATLAS publishes T1611 as the ATT&CK reference for this technique.
Cross-system agent lateral movement and execution boundaries →
Attack scenarios using this technique
References
MITRE ATT&CK® and ATT&CK® are registered trademarks of The MITRE Corporation. ATT&CK data © 2026 The MITRE Corporation. This work is reproduced and distributed with the permission of The MITRE Corporation (ATT&CK v20.1). MITRE ATLAS™ data (v2026.10) © MITRE, used under the Apache License 2.0. Securitain is not affiliated with or endorsed by MITRE. Explanations, scenarios and control mappings on this page are Securitain’s own analysis. Sources and licensing